Is Cosmetic Science Really "Bad"? Part IV: How Scientific is Cosmetic Science?

The third and most recent column in this series investigating whether cosmetic science is "bad" produced a number of reactions. Of course cosmetic science is not "bad" but it can be done in a bad way. Therefore, the initial question of “Is cosmetic science really 'bad'?” changes to: “Are you doing ‘bad’ cosmetic science?” It becomes personal to suggest that someone is doing bad science, which was obvious from the emotional reactions.

After part III of this series, a discussion followed on the Society of Cosmetic Chemists' LinkedIn group, where many people said, "It’s not me, but ..." filling in the blank with their favorite group of culprits such as marketing, consumers, competitors, etc. While these individuals are correct in stating that it is not only cosmetic scientists that may get it wrong, still, cosmetic scientists are not correct "beyond any reasonable doubt." While this was acknowledged by most, the attitude in most reactions was that the other party was more guilty and should change (first). But why should small bankers hand in their small bonuses when big bankers continue to cash big bonuses? Why should someone care about the environment if that company next door is not as environmentally friendly as they could be?

As my mother used to say (and I'll bet your mother, too): "Changing the world means you have to start by changing yourself." But  it is not easy to implement change in a society, as Barack Obama is finding out the hard way. So why would it be easy to implement change in the society of cosmetics? The point of this series of columns is to urge those scientists who find the reputation surrounding cosmetic science to be unjust to change, although change requires action and energy and does not happen by itself. But let’s return to cosmetic science to continue the evaluation of how good (or bad) it is.

Six questions from Michael Shermer’s Baloney Detection Kit have been discussed in this series. The results were not flattering for cosmetic science (read: cosmetic scientists) but cosmetic science did not score badly on all points. Questions seven and eight of the Baloney Detection Kit discriminate true science from borderland science and non-science or nonsense.

Question seven asks, “Is the claimant employing the accepted rules of reason and tools of research, or have these been abandoned in favor of others that lead to the desired conclusion?” I would not deliberate on a paper I once edited for the magazine of the International Federation of Societies of Cosmetic Chemists (IFSCC). A cosmetic scientist wrote in a submitted paper that something was significant at the p = 0.5 level, leading me to ask whether this was a typo and this should instead read p = 0.05. The author's response was no, and that their statistical analysis showed the result was statistically significant at the p = 0.5 level, which is as significant as flipping a coin ... but the author wanted this published.

Needless to say, the paper was rejected. A cosmetic scientist may be factually right but it is against the accepted rules of science and tools of research to call anything with a p value above 0.05 significant. This individual does not represent the behavior of the whole cosmetic industry but it serves as a reminder for papers claiming statistical significance at p < 0.1, or p = 0.07. Even in one of my own papers, I have stated that my results achieved a p value of 0.055 but I always add that that is not significant. Occasionally, cosmetic scientists do not employ the accepted rules of reason and tools of research but this occurs infrequently; rather, almost the opposite regularly happens in our industry.

In relation, the cosmetic scientist also does his or her utmost to portray a scientific image. For example, a number of articles in trade magazines reference Nature and Science, two high impact scientific journals that do not in fact deal with cosmetics. These references are intended to give the impression that the article is scientific.

Another example of (sometimes not) “employing the accepted ... tools of research” is the use of questionnaires in clinical testing. There is a striking difference between the methods for conducting cosmetic claim substantiation in Europe and Asia compared with the United States. Wherever possible, the methodology will use skin bioengineering equipment in Europe and Asia, whereas questionnaires are used in the United States when possible. Europe and Asia may use questionnaires as supporting evidence whereas the United States may use skin bioengineering equipment as its supporting evidence.

Cosmetic scientists know that questionnaires can be leading, sometimes even misleading. If a woman is told that she is testing anti-wrinkle products, nearly any product will work (to some extent). Referring to Perry Romanowski’s comment during the LinkedIn discussion, this is what I would like to call the "gullible consumer" effect. Some consumers believe anything because they want to believe everything. Of course, questionnaires can be constructed in good and bad ways but measuring the objective nature of skin via bioengineering equipment is preferred. Everybody will accept that an instrument is more objective than asking a volunteer whether she felt better after having used the product. But to quote the late, great Professor Albert Kligman, who sadly died last month, "A fool with a tool is still a fool!"

Still, skin bioengineering studies can go wrong and both questionnaires and skin bioengineering equipment must be used properly, and I have no reason to assume that this is deliberately not the case. Questionnaires will, however, always be deemed as less objective unless one can see the exact questionnaire used and whether repeat questions asked in different ways resulted in the same result.

To state that the cosmetics industry is consistently abandoning the rules of science to lead to the desired outcome is simply inappropriate. However, the way in which a cosmetic scientist conducts research can be scientifically unsound because the choice is made to not investigate some things (or not report them). Nobody likes to tell the boss that something did not work. Any scientist knows how data can be "massaged" not only in the cosmetic industry but everywhere. All scientists will always aim to portray themselves or their data in the best possible way but they cross the line when they make things up or deny things when asked. I have sufficient evidence of the cosmetics industry not telling the other side of the coin but no evidence whatsoever of consistently abandoning the rules of science to lead to the desired outcome.

The latter relates more to the eighth question of the Baloney Detection Kit, which asks: “Has the claimant provided a different explanation for the observed phenomena, or is it strictly a process of denying the existing explanation?” Since cosmetic scientists are all confirmation-biased to some extent, this is an area where they can easily go wrong. Especially because cosmetic science is a commercial science. Why? Academic science is conducted to find out how things work and to test or deny a hypothesis, whereas commercial science is conducted to find reasons for selling a product, mostly to discover how good a thing is relative to nothing or benchmarks. A large portion of the cosmetics industry is conducting science to sell products and not to explain things. Therefore, cosmetic science is often comparative science (how good is my product or my ingredient doing relative towards my competitor’s product or ingredient?) instead of explanatory science (how does my product work?).

Manufacturers are selling action and not mechanisms of action. Ingredient suppliers are selling mechanisms of action as they are part of the marketing concept. Ingredient suppliers, therefore, offer more explanation than the manufacturing industry. So again, the cosmetics industry is doing OK here. Not better than the rest of the industry but simply in line with the rest because cosmetic scientists normally do not offer explanations as to how their products work,

Sometimes, it is even the reverse. Cosmetic scientists offer the explanation (currently by using genomics and proteomics) without offering the observed phenomena. They then sell the explanation based on in vitro evidence of up- and down-regulation of various enzyme systems without providing the in vivo evidence for the efficacy of the product. Reasons given include the fact that biopsies need to be taken on the face, which is not ethically acceptable. That does limit the commercial value of the product but it does not devalue the cosmetic science of such products.

Occasionally the denial of an existing explanation occurs in cosmetic science. An old-fashioned hygrometer consists of a hair from the tail of a horse fixed in a spring mechanism. Based on the water uptake from the environment, the length of the hair will vary, which indicates the ambient relative humidity on a scale. If a hygrometer works based on the water uptake of a hair from the environment, why would the cosmetic scientist need a concept like hair moisturization? Hair is supposed to change with its environment, not resist it. If the cosmetic scientist uses hair in hygrometers to measure the relative humidity, why would hair have systems to retain water? This does not take away the fact that hair in a dry environment is brittle and that one would like to do something about it.

Another example is the green movement. While there is nothing wrong with this concept in cosmetics, it is "bad" science to state that molecules are better because they are natural. The human body cannot check the carbon footprint of a glycerin molecule. It will react to every CH2OH-CHOH-CH2OH molecule in exactly the same way.

In summary, I think that although individual examples can be found where the cosmetic scientist does not fully adhere to the principles outlined in the Baloney Detection Kit, as a whole, we are OK on this eighth question.

Eight of the ten questions have now been addressed. This part of the series did not make cosmetic science sound "bad" but it also makes this column less provocative than the previous one. However, the worst of the questions is still to come. Be prepared.

More in Method/Process